INVITED: Design of Complex Oxides aided by Ab-initio and Data Science Tools

Structural and chemically complex alloys have many engineering applications. Here I would like to present two examples from my recent research on designs of complex oxides aided by ab-initio and data science tools. The first example is to study the thermodynamic stability of defects in alumina at the oxide-electrolyte interfaces. Alumina can act as passive oxides to enable high corrosion resistances for alloys. However, exposure to certain chemical species such as chlorine can result in the breakdown of passive oxide layers and localized corrosions. Due to the complex electrochemical interfacial structures, the intrinsic mechanisms for chlorine effects on localized corrosion are still unclear. Our ab-initio molecular statics and dynamics studies reveal the intrinsic mechanisms for chlorine effects on defect stabilities of alumina at the alumina-water interface under different local chemical and structural conditions, which can be helpful for the design of passive oxide layers with high resistance to localized corrosions. The second example is to use machine learning to predict density and elastic moduli of SiO₂-based glasses. Multicomponent SiO₂-based glasses with high moduli and low densities are ideal structural components in composite materials. However, it is difficult to find a universal expression to predict the moduli and densities according to the glass composition due to the complex glass structures. Our machine learning approach relies on a training set generated by high-throughput atomistic simulations and a set of elaborately constructed descriptors with the fundamental physics of interatomic bonding. By just training with a dataset only composed of binary and ternary glass samples, our model shows very promising capabilities to predict the density and elastic moduli for k-nary SiO₂-based glasses beyond the training set. The predictions of our model are comprehensively compared and validated with a large amount of both simulation and experimental data.